Here’s Help For A Huge Hardship
April 16, 2014 Ted Holt
| 
 
 
 Multimillion-row tables are more and more common in IBM i shops these days. Querying those monsters can be a strain on the system. Fortunately, IBM gave us some help in DB2 for i 7.1. The help comes in the form of a new wrinkle in indexing. The smart people who produce the world’s greatest relational database management system found a way to store aggregate (summary) information in encoded vector indexes. Here’s an example. Assume a table (physical file) of sales history, such as this one: 
create table Sales
  ( ID                integer as identity,
    Invoice           dec  (9,0),
    Line              dec  (3,0),
    InvoiceDate        | 

 
							  
					
